174 ALASKAN GLACIER STUDIES In other parts of the St. Ellas Range, geologists, army officers, prospectors, and others saw and heard many great avalanches during the earthquakes, although some of these observers were several hundred miles from the earthquake centre; but in these remote regions the avalanches were not in such abundance as near Yakutat Bay. One who has read Russell's vivid descriptions of the great avalanches in the valley of upper Newton glacierl in 1890 and 1891, some of which advanced a half mile across the glacier, or who has seen Bella's wonderful photographs2 of the avalanche-scarred slopes near the heads of the glaciers of Mt, St. Elias in 1897, in both cases representing avalanching under normal conditions, not under the exceptional conditions of three weeks of earthquake shocks, and who has noted the enormous area of snow slopes in unstable equilibrium near Yakutat Bay, could have no serious doubt of the tremendous accessions of snow and ice that might have been received by many of the Yakutat Bay glaciers during September, 1899. Upon these evidences, we are convinced that there can be no question but that there was, in that year unusual addition to the reservoirs of the glaciers of Yakutat Bay and vicinity. The mountains here are lofty, with steep slopes, and the snowfall is very heavy. The same statement applies to other parts of the mountains of the Alaskan coast region, it is true; but, although these other sections have experienced earthquakes, none have suffered such severe and continued shaking since Alaska began to be settled by white men as the Yakutat Bay region was subjected to in September, 1899. That period of shaking was notable for its intensity and duration, even in a region of frequent and severe earthquakes. If the avalanching of snow and ice to glacier reservoirs is competent to cause an advance in the glacier termini, the advance resulting from that particular period of shaking should be notable for its extent and intensity. Lesser shaking might well cause slower and less spectacular advances,' of which the slighter forward movements of the glaciers of Alaska and other regions may be instances; but vigorous shaking, continued through three weeks, may well produce additions to the glacier supply that cannot be disposed of in so simple a manner. Until the studies of the Yakutat Bay glaciers in 1906 the theory of advance of glaciers in response to earthquake shaking had not seemed called for in explanation of the phenomena of glacier fluctuations, Prom the studies of that year it became evident that some unusual cause was necessary to explain the remarkable changes observed in the glaciers, and the earthquake theory was therefore proposed. Our later studies have discovered new facts in support of this theory and none opposed to it. In view of the newness of this theory, and of its possible application to glacier advances in other regions, we deem it worth while to state it fully, to give a complete statement of the reasons why it seems to be necessary, and to state the facts which support it. Failure of Other Theories. One of the main reasons for proposing the new theory of advance due to an impulse derived from earthquake shaking is the failure of all other explanations suggested. It was evident that we had here such a strong expression of some cause for glacier advance that we were warranted in considering not only current theories, but unusual and even apparently improbable theories, and in inventing for consideration any that might by any possibility explain the phenomena. Little or no help could be obtained from observations in other regions, because the phenomena i Russell, I. C., Nat. Geog. Mag., Vol. m, 1891, pp. 155-156; 18th Ann. Kept, U. S. Geol. Survey, 1892, p. 50_ * The Ascent of Mt. St. Elias, London, 1900, photograph facing p. 180, and many others.